Surgery is being used increasingly as a treatment for medically intractable epilepsy, and anterior temporal lobectomy (ATL) is the most common surgical procedure. However, there are very few controlled or prospective studies of the tests employed to localize a resectable epileptogenic region or of comprehensive outcome from ATL. The long-term objectives of this study are to develop a strategy that will identify by noninvasive tests the largest possible number of patients with intractable epilepsy who can be successfully treated with ATL and determine comprehensive outcome from ATL using multiple measures. The expected results will decrease the number of patients who require long-term intracranial electroencephalography (EEG) prior to surgery and will provide critical information for a more complete and appropriate presurgical informed consent process. The specific aims of the study are to; 21) determine an accurate, efficient, and cost-effective algorithm of noninvasive tests that selects a majority of patients with intractable epilepsy who can be treated successfully by ATL, 2) determine the outcome from ATL using multiple measures of function, quality of life, and seizures, and 3) determine the morbidity and mortality from ATL. The study is a prospective, nonrandomized controlled trail divided into two phases; phase I tests the noninvasive algorithm and phase II determines comprehensive outcome. The predictor variables in the algorithm are based on interictal EEG, electrographic and clinical seizures analyzed by video/EEG monitoring, neurological examination, history of febrile convulsions, hippocampal abnormality on magnetic resonance imaging (MRI), and an increased temporal creatine/N-acetyl aspartate ration on MR spectroscopy. Combinations of weighted test criteria will determine candidacy for ATL without requiring intracranial monitoring conformation of the epileptogenic zone. The outcome measure for phase I is complete cessation of seizures. The predictor variable for phase II is ATL. The outcome measures for phase II are seizure cessation, quality of life, neuropsychological function, psychiatric status, vocational and educational achievement, and patient satisfaction. Quality of life, neuropsychological function, and psychiatric status will be assessed by previously standardized and validated tests. All outcome measures will be repeated yearly on each subject until the end of the study; the enrollment period will be four years so that all patients will have a minimum of 1 year follow-up. Seventy-five percent of subjects will be followed for more two years. Controls will be patients with temporal lobe epilepsy confirmed by the algorithm who defer surgery or are ultimately not candidates for resective surgery (i.e., bitemporal seizure onset) and are followed for a minimum of 1 year. Statistical analysis of phase I will include logistic regression analysis, proportional hazards model for time to first seizure, and a step-wise procedure. Phase II statistical analysis will include as appropriate a categorical format (chi square or Fisher's Exact test), repeated measures analysis, interaction assessment, and ANOVA with covariance adjustment.